search

carstenbauer / MonteCarlo.jl

Classical and quantum Monte Carlo simulations in Julia
https://carstenbauer.github.io/MonteCarlo.jl/dev/
Other
185 stars 19 forks source link

Document observables for Hubbard DQMC. #46

Open MasonProtter opened 4 years ago

MasonProtter commented 4 years ago

Currently the observables are :conf, :Greens and :BosonEnergy. We should document what these are and what the actual data they contain means. In particular, I'm not sure how to interpret the matrix returned by e.g.

julia> meas[:Greens].obs |> mean
8×8 Array{Float64,2}:
  0.486632   -0.307531    -0.0152945   0.0458478   0.0299359    0.0384947  -0.0139994  -0.25898  
 -0.310227    0.538009    -0.304965    0.0196503   0.049056    -0.0135113   0.0716403   0.0193743
 -0.0263509  -0.269518     0.518684   -0.237224   -0.0447043    0.0557874   0.0241328   0.0390698
  0.0654728   0.00375744  -0.317623    0.409047   -0.331587     0.0385827   0.042445   -0.0055775
 -0.018052    0.0400457    0.0192787  -0.237427    0.605781    -0.309867    0.0151516   0.0356679
  0.032978    0.00207081   0.0471798   0.0185239  -0.247373     0.357371   -0.261951    0.0109985
  0.0188517   0.0305705   -0.0226769   0.041055    0.00292663  -0.342347    0.561436   -0.235713 
 -0.335868    0.0088788    0.0530074  -0.0274625   0.0661984    0.0356364  -0.361961    0.438607

I suppose that one dimension is the imaginary time and another is the spatial coordinate, but I'm not sure which is which.

carstenbauer commented 4 years ago

I suppose that one dimension is the imaginary time and another is the spatial coordinate, but I'm not sure which is which.

No, this is the equal-(imaginary) time Green's function (see the outdated https://crstnbr.github.io/MonteCarlo.jl/latest/flavors/dqmc/). It has size (flv * N, flv * N) where N is the number of lattice sites and flv is the number of fermion flavors (spin, band, whatever).

Should definitely be documented.